Professional Documents
Culture Documents
Sanderset Al 2000 PDF
Sanderset Al 2000 PDF
ABSTRACT: Most unpaved road dust measuring devices have employed one or more of the air sampling
techniques used by atmospheric scientists, namely, sedimentation, filtration, and photometric techniques. The
dust measurement quantifications have been for a section of unpaved road either involving the use of moving
devices, or at a point(s) along a section involving the use of a stationary device. Devices based on the photometric
technique have measured and quantified dust on a real-time basis without laboratory analyses of field data, while
devices based on the sedimentation and filtration techniques have relied on laboratory analyses of field data for
dust quantification. The Colorado State University dustometer, an experimental road dust measurement device
based on the filtration technique is presented in this paper. The device and technique were developed, field
tested, and used in comparative fugitive dust emission studies at Colorado State University and documented by
Addo and Sanders. The device, which is basically a moving dust sampler, consists of a standard high volumetric
sampler with associated filter media and accessories mounted on the rear of a pickup truck. Many dust mea-
surements can be made in a single day using the dustometer. The dustometer has proved to be a very quantitative,
reproducible, and precise measurement device for unpaved road dust measurements.
Downloaded 25 May 2012 to 137.121.21.31. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
technique in remote locations is impossible without a portable Studies by Iowa State University
generator(s).
The photometric technique is based upon light scattering or In an effort to quantify the dust fallout from aggregate sur-
absorption properties of particles passing through a light faced roads, dust particulate collections were conducted by
source. Particles of all sizes may reduce visibility by absorbing Iowa State University during the spring and fall months of
light. The amount of light scattered depends upon the concen- 1978 through 1980 in a dust control and unpaved road surface
tration, size, refractive index, shape, and color of the sus- improvement project (Hoover et al. 1981). The dust collection
pended particles. was done in order to assess the effectiveness of the various
A particle size of 10 m (PM10) or less are most often mea- dust control and surface improvement processes. The dust col-
sured because of the health hazard posed by such micrometer lectors consisted of a 15.2 cm (6 in.) diameter ⫻ 19 cm (7
size particulate matter in the atmosphere. Particulate matter of in.) deep semirigid plastic container, mounted 100 cm (3 ft)
size 10 m (PM10) or less are noted to be retained in the above ground level in accordance with ASTM D 1739. The
human respiratory system (K. Gottschalk, unpublished inde- dust data were analyzed by plotting on log-log scale the quan-
pendent study, 1994). tity of the dust deposited in pounds per acre per day per 100
The need to quantify unpaved road dust generated by ve- vehicles versus the distance from the centerline of the road-
hicles has led to many quantitative dust studies. Several of way.
these studies have proposed mathematical models for the gen- Kittleman (1973), Eaton (1964), and A. J. Lutenegger (un-
eration and distribution of particulate emission from various published PhD dissertation 1979) concluded that the deposi-
sources (Robinson et al., 1971; D. C. Becker, unpublished MS tion of particulates due to sedimentation varied logarithmically
thesis, 1978). D. C. Becker (unpublished MS thesis, 1978) with distance from the source of generation. However, during
applied a Gaussian plane model to project dust generation the Iowa project, Hoover (1981) observed that the dust fall in
from unpaved roads in Iowa and Kansas and concluded that any of the directions measured did not follow a perfect log-
dust levels of 60 mg/m3 of air drawn using a high volumetric log relationship. He noted that dust quantities during any pe-
sampler were not exceeded at approximately 20 ft from either riod of the measurement were affected by the direction and
side of the unpaved roads. Hoover et al. (1973) using station- velocity of the prevailing wind, topography of the road, trees,
ary collectors placed adjacent to unpaved roads measured dust vegetation, moisture content of the roadway surface, type of
levels in the range of 6–110 lbs/acre/day. The U.S. Environ- roadway surfacing, humidity, and other factors.
mental Protection Agency (EPA) in its attempt to estimate the Dust measurement by ASTM D 1739 is not without prob-
amount of dust generated from unpaved roads published a for- lems. In addition to the time-consuming laboratory work of
mula [AP-42 (Compilation 1988)] that can be used to estimate sample drying, removal of foreign matter and weighing of dust
unpaved road dust amounts. The empirical equation is based samples, the amount of dust collected can only quantify the
upon the properties of the road base material; the number, road dustiness at a specific point along the stretch of road.
weight, and speed of passing vehicles; and the number of dry Due to the natural variations on the roadway (upgrades, down-
days per year. The formula predicts the emissions of particle grades, curves, etc.), the level of dust measured at a single
size in the range of 10 m (PM10) or less (Compilation 1988). point cannot be used to characterize the degree of dustiness
The equation is defined as follows: for an entire road section.
Downloaded 25 May 2012 to 137.121.21.31. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
ter) was used to measure the dust density (i.e., air opacity). The RDM measures the opacity of the air as the test truck
The device consisted of a 12-V, 100-W aircraft landing light travels the unpaved road. A typical instrument response indi-
used for illuminating the dust cloud produced by the passing cates that the measured air opacity varies as the vehicle moves
vehicles and a recorder that measured the percent of light along the roadway. The variations are explained to be due to
transmitted through the dust cloud. Dust fallout was also col- spotted sunlight, which gave a higher reading than the shaded
lected using buckets in accordance with ASTM D 1739. The spots. Higher readings were measured on curves and lower
setup of the test is shown in Fig. 1. Flash boxes were used to readings on upgrades. These inherent road variations, along
determine vehicular speed. Many factors influenced the pro- with changes in the types of surface material according to
cedure, including temperature, humidity, sunlight, wind speed, Irwin et al. (1986) make it difficult to use a point response as
and direction. Although the photoelectric device (densitome- a valid measure of road dust levels. As a result of the variation
ter) quantified the dust measurement directly in the field with- in road dust levels, a section descriptor rather than a point
out any laboratory measurement, it provided a measurement descriptor is used to quantify the road dustiness in the RDM
valid for only a specific point along the road section. The study procedure.
concluded that the amount of dust generated by a vehicle tra- The major advantage of this method is the absence of lab-
versing a gravel road is related to the vehicle’s speed, gross oratory analyses of the dust measurements. The data are gen-
weight, the number of wheels, and the load shape factor that erated in real time; however, a problem with the RDM tech-
influences the degree of turbulence created by the vehicle. The nique for measuring road dust is that the infrared transducer
study also showed that there was good correlation between the and the sensor mounted in the metal duct needed frequent
dust density measurement and the dust fall measurement. cleaning and maintenance because they were always coated
with dust, thus affecting the accuracy of the dust measure-
Dust Volume Sampler ments.
In a USDA Forest Service dust abatement study, Langon
(1984) measured dust over a section of roadway instead of at Summary of Previous Work
a particular point along the section. His explanation for using
A review of the dust measurement or sampling devices in-
a section descriptor instead of a point descriptor in the dust
dicates the value of the section measurement device versus a
quantification was that the characteristics of a road section,
point measurement device. The former allows the integrated
such as upgrades, downgrades, curves, variation in road sur-
analysis of dust production on the entire road section, whereas
face material, moisture content, shading of the roadway by
the latter is indicative of only one point along a road section;
trees, etc., influenced the variability of dust generation from
therefore, the dust data are more susceptible to errors intro-
the section of roadway that cannot be quantified adequately
duced by changing wind directions and vehicle speed as well
with buckets. Using the filtration sampling technique Langon
as the number of vehicles. The section measurement devices
put together a device comprising a cyclone dust collector
allows for the collection of a large amount of data in a single
mounted behind the left rear wheel of a 3/4-ton pickup truck.
day, whereas a point descriptor device generally generates one
As the test truck drives along the test section, dust kicked up
data point in a month. The light scattering (photometric) tech-
by the left wheel enters the cyclone dust collector, and at the
niques, when mounted on a vehicle, produced section mea-
end of the run the dust collected is weighed. Driving the truck
surements on real-time bases without any laboratory analyses.
at a constant speed of 25 mi/h, the amount of dust collected
The main issues with this measuring device are that it is rel-
ranged from 0 to 2.5 g/5,000-ft of roadway. A major problem
atively expensive to build, and it has a calibration problem
using this device was the amount of time between test runs.
because of a buildup of dust on the sensors.
The time to remove the sampler and prepare the sampler for
another test run allowed only a limited number of test runs in
a day. COLORADO STATE UNIVERSITY DUSTOMETER—
AN EXPERIMENTAL ROAD DUST
Road Dust Monitor (RDM) MEASUREMENT DEVICE
In a cooperative study by the Cornell University Local Between the period of 1992–1995, a Mountain Plains Con-
Roads Program and the USDA Forest Service, Irwin et al. sortium and department of transportation sponsored research
(1986) developed and field tested a device based on the pho- project entitled ‘‘Relative Effectiveness and Environmental
tometric principle called the RDM. The idea behind the RDM Impact of Road Dust Suppressants’’ was undertaken in the
was to provide a standard, repeatable, and quantitative method Civil Engineering Department at Colorado State University
of monitoring dust production from unpaved roads. One of the (Sanders et al. 1997). One of the objectives of the project was
objectives of the project was to provide USDA Forest Service to develop an inexpensive dust measuring device and a tech-
road managers with a more reliable and consistent approach nique that could be used to evaluate the performance of dif-
to deciding when to maintain an unpaved road, rather than the ferent dust suppressing chemicals. This part of the paper there-
qualitative method, which is based entirely on human judg- fore describes the dust measuring device and the technique
ment and experience. The RDM consists of a 20 cm (8 in.) ⫻ employed to assess the relative effectiveness of the different
30 cm (12 in.) galvanized sheet metal duct and an instrument dust suppressants, namely, calcium chloride, lignosulfate, and
package. The duct, which has a transducer system mounted in magnesium chloride. In all, four 1-mi test sections were built:
it, is positioned behind the rear wheel of a 1/2-ton pickup three treated, and one untreated which served as the control.
truck. Early attempts to measure the dust fallout from the test sec-
Dust kicked up by the truck as it moves along the roadway tions using the bucket method in accordance with ASTM D
enters and passes through the duct. The transducer contains a 1739 in the summer and fall of 1992 proved ineffective and
light source and a photoelectric sensor. The light beam shines inefficient. The reasons were that (1) it was difficult to obtain
onto a reflector, and the reflected light is detected by the sen- permission from the landowners to set up the dust collectors;
sor. Therefore, dust passing through the duct scatters or ab- (2) the land along the test sections are grazing pastures, and
sorbs the light beam reducing the amount of light reflected there was the problem of livestock continually tampering with
back to the sensor. The sensor measures the reflected light and the dust collectors; and (3) the method requires too many ve-
converts it to a voltage that is displayed on the indicating hicles to drive by the buckets before a substantial amount of
meter position in the cab of the truck. dust can be collected for quantification. The prevailing wind
532 / JOURNAL OF TRANSPORTATION ENGINEERING / NOVEMBER/DECEMBER 2000
Downloaded 25 May 2012 to 137.121.21.31. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
FIG. 2. Half-Ton Truck with Attached Dustometer Used in Tests
Downloaded 25 May 2012 to 137.121.21.31. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
FIG. 5. Dustometer Opened Showing Location of Filter Paper
FIG. 6. (a) Dustometer Mounted on Back Bumper of Test Truck; (b) Side View of Test Truck Showing Flexible Tube Connecting Suc-
tion Pump to Dustometer
signed to allow for easy removal and replacement of the filter TABLE 1. Typical Dust Measurement Data (Speed: 45 mi/h;
paper. Length of Run: 1-mi; Test Section: Untreated)
To run a typical dust measurement test, the Colorado State Weight of dust
University dustometer is set up as shown in Fig. 6 with the Sample number (g)
flexible hose connecting the suction pump to the filter. Once (1) (2)
the generator is started, the pump can be turned on and off 1 2.85
from inside the cab of the truck making the dust measurement 2 2.60
completely a one person operation. For each test run the gen- 3 2.83
erator is started, and the truck is driven at a stipulated speed. 4 2.86
At the beginning of the 1-mi mark when the truck is moving 5 2.87
at the desired speed, the pump is turned on and a portion of 6 2.47
7 2.62
the dust kicked up by the rear left tire is drawn into the filter. 8 2.48
At the end of the 1-mi run the pump is turned off, and the 9 3.09
vehicle is then brought to a stop. Three runs were made in the
Note: Mean = 2.74 g; standard deviation = 0.21 g; variance = 0.04 g.
same driving lane for each of the four test sections in a single
day. The dust-laden filters were carefully removed from the
filter box after every run, placed in preweighed plastic bags, As indicated by the data and its analysis, the dustometer is
taken to the laboratory, and weighed. The average of the three quite precise for a field measurement device.
measurements for each test section were compared to assess During the initial testing of the dustometer it became quite
the relative effectiveness of the different treatments. obvious that the speed of a vehicle is a very important factor
To demonstrate the precision of the dustometer as an ex- in dust generation from unpaved roads. The faster the vehicle
perimental road dust measurement device, nine replicate mea- travels, the more dust is generated. As a result, speed control
surements were taken on the 1-mi untreated test section at a is one of the techniques employed for dust control on unpaved
speed of 45 mi/h. Table 1 shows the data and its distribution. gravel roads. To help quantify this phenomenon, dust mea-
A mean of 2.74 g was obtained with a standard deviation of surements at various speeds were made using the dustometer.
0.21, a variance of 0.04, and a coefficient of variation of 7%. Fig. 7 shows the amount of dust sampled at speeds of 20, 30,
534 / JOURNAL OF TRANSPORTATION ENGINEERING / NOVEMBER/DECEMBER 2000
Downloaded 25 May 2012 to 137.121.21.31. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
Additional dust data were collected later for speeds of 25,
35, and 45 mi/h to see if the relationship remains linear. Add-
ing these data points to the plot did not change the linear
relationship as shown in Fig. 9. The coefficient of linear re-
gression r 2 for the plot is 0.95. Approximately 2.3 g of dust
were measured at a speed of 20 mi/h, and about 7.3 g were
measured at 50 mi/h indicating a more than threefold increase
in dust generation.
CONCLUSIONS
The sedimentation, filtration, and photometric techniques of
air sampling have been the techniques employed by nearly all
FIG. 7. Graph of Weight of Dust Generated versus Vehicle attempts of measuring unpaved road dust. The resulting dust
Speed for 1-mi Run measuring devices developed have been either a stationary de-
vice, which measured road dust at a particular point, or a mov-
ing device, which measured road dust over a section of road.
The Colorado State University dustometer, which employs the
filtration technique, improves on the existing moving devices.
Although the dustometer technique involves weighing of filters
in the laboratory, the technique is still fast, and it enables large
amounts of precise dust data to be gathered in a short period
of time. The dustometer measures directly the dust mass,
whereas other moving devices measure dust indirectly through
air opacity. The dustometer is simple to build, easy to operate,
and very inexpensive (less than $1,000.00/device).
APPENDIX. REFERENCES
FIG. 8. Graph of Weight of Dust Generated versus Vehicle Addo, J. Q., and Sanders, T. G. (1995). ‘‘Effectiveness and environmental
Speed for Run Time of 3 min impact of road dust suppressants.’’ MPC Rep. No. 92-28A, Mountain
Plains Consortium.
ASTM. (1993). ‘‘Standard method for collection and analysis of dustfall.’’
D 1739-98, West Conshohocken, Pa.
Compilation of Air Pollutant Emission Factors. (1988). 4th Ed., EPA
Publ. AP-42, including Supplements A & B, U.S. EPA.
Eaton, R. A. (1964). ‘‘Windborne volcanic ash: A possible index to polar
wandering.’’ J. Geol., 72(1), 1–35.
Federal Highway Administration (FHWA). (1992). Local low volume
roads and streets, U.S. DOT, Washington, D.C.
Handy, R. L., Hoover, J. M., Bergerson, K. T., and Fox, D. E. (1975).
‘‘Unpaved roads as sources for fugitive dust.’’ Transp. Res. News, 60,
6–9.
Hoover, J. M., Bergerso, K. L., Fox, D. E., Denny, C. K., and Handy, R.
L. (1973). ‘‘Surface improvement and dust palliation of unpaved sec-
ondary roads and streets.’’ Final Rep. HR-151, Iowa Highway Research
FIG. 9. Graph of Weight of Dust Generated versus Vehicle Board Project.
Speed Using All of Data for Run Time of 3 min Hoover, J. M., Fox, D. E., Lustig, M. T., and Pitt, J. M. (1984). ‘‘Mission-
oriented dust control and surface improvement processes for unpaved
roads.’’ Final Rep. H-194, Iowa Highway Research Board Project.
40, and 50 mi/h from the untreated test section over a 1-mi Irwin, L. H., Taylor, D. J., and Aneshansley, D. J. (1986). ‘‘Device to
run. Because the dustometer dust measurement involves the measure road dustiness on aggregate surface roads.’’ Cornell Local Rd.
suction of dust as it is generated, the volume of the dust cloud Program Rep. 86.5, Dept. of Agric. Engrg., Cornell University, Ithaca,
drawn is related to how long the suction pump is allowed to N.Y.
Langdon, B. B. (1984). ‘‘Dust volume sampler documentation.’’ USDA
run. This means that for the 1-mi run, the weight of dust drawn Forest Service, Mount Hood National Forest.
will be different at each test period. To make the data in Fig. Robinson, E., and Robbins, R. C. (1971). ‘‘Emissions, concentrations and
7 comparable, the amounts of dust were adjusted for a 3-min rate of particulate atmospheric pollutants.’’ Rep. 4076, American Pe-
run to make the dust collection time equal for all the speeds. troleum Institute, Washington, D.C.
Fig. 8 shows the adjusted data and indicates a linear relation- Sanders, T. G., Addo, J. Q., Ariniello, A., and Heiden, W. F. (1997).
ship between the speed and amount of dust generated. The ‘‘Relative effectiveness of road dust suppressants.’’ J. Transp. Engrg.,
ASCE, 123(5), 393–397.
coefficient of linear regression r 2 for this plot is 0.98. The fact Wellman, E. A., and Barraclough, S. (1972). ‘‘Establishment of acceptable
that Fig. 8 appeared linear was surprising as visual inspection dusting criteria for aggregate surface roads.’’ Study 7110, USDA Forest
and subsequent videotaping appeared to indicate a nonlinear Service Admin., Winema National Forest.
relationship. What the Forest Service does. (1983). USDA Forest Service.
Downloaded 25 May 2012 to 137.121.21.31. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org